3.3. GLASS MEMBRANE SENSORS 1. History of the Development of a Glass Sensor of pH
3.3.2. Glass Structure and Sensor Properties
How does a glass membrane develop a signal that is sensitive to pH? The answer lies in the structure of the glass. The atomic structure of glass includes negatively charged sites that bind H+ ions. An early pH sensitive glass made by Corning has the composition of 22%
Na2O, 6% CaO, and 72% SiO2[15]. The term “glass” suggests that the material is amor- phous; there is no long-range crystalline order to the arrangement of atoms. It is helpful
k k
+ +
+ + +
+
+
+ +
+ +
Si4+
O
O
(a) (b)
++
+ ++
= Silicon = Oxygen = Hydrogen = Sodium ion = Calcium ion (2–)
(2–) (2–)
(2–)
–
– – –
– – –
– –
–
– –
FIGURE 3.9 Glass structure. (a) Silicate tetrahedron. (b) Conceptual model for glass showing tangled strands made from chains of silicate units. Charges remaining on some silicate units are balanced by cations. Occasionally, hydrogen ions bond to anion sites. Source: Adapted with permission from Perley [18]. Copyright 1949, American Chemical Society.
to think of the framework as an assembly of silicate units that form tetrahedra. A silicon (4+) cation sits in the center of each tetrahedron (Figure 3.9a). An oxygen (2−) ion sits at each vertex of the tetrahedron. In a perfect network solid, each of these oxygen ions would also be a part of another adjacent tetrahedron as well. One can think of the two negative charges associated with each oxygen atom being divided between two tetrahedra. In that sense, each oxygen ion contributes half of its charge to each tetrahedron it participates in. As a result, each individual tetrahedron would be balanced with respect to charge (4+
for the silicon atom+[4 oxygen atoms]×[−1]). Although there are possibly small regions where this uniform Si–O network extends in all directions, the structure mostly resem- bles chains of tetrahedra, tangled together with frequent cross-links between chains (as in Figure 3.9b). Any oxygen atom that does not bridge between two silicon atoms confers a net negative charge on the one tetrahedron that it belongs to. These charges attract Na+ or other cations as the original molten mixture cools in the glass-making process. Upon exposure to water, hydration of the surface of the glass occurs to a depth of about 0.1μm.
As water penetrates this thin zone, sodium cations can be displaced by H+ions.
Na+(glass)+H+(solution)⇌Na+(solution)+H+(glass) (3.19) The sites in the structure where this exchange occurs are the charged silicate tetrahe- dra. The equilibrium constant for this exchange strongly favors the hydrogen binding to the glass. George Eisenman rationalized this preference for H+as being the result of the
k k higher field strength at the surface of the H+ion because of its smaller ionic radius [19].
The positive charge is spread out over a smaller surface of the hydrogen cation making the intensity of its field greater at the hydrogen ion surface than for the larger sodium ion sur- face. It is also helpful to think of this in terms of functional groups. Protonating a silicate anion creates a silanol group – a weak acid.
≡Si−O−(membrane)+H+(aq)⇌≡Si−OH(membrane) (3.20) Within the hydrated zone of the glass wall, there can exist some H+ ions in solution.
These H+ions balance the total charge on the glass surface, but while in solution, these ions are on the other side of the electrical double layer formed at the membrane surface.
They are part of the charge that creates an electrochemical potential energy difference at the interface. The extent to which the charge appears on the membrane depends on the degree of dissociation of the silanol groups on the surface. If a silanol group dissociates, the H+ becomes solvated and becomes part of the solution side of the interface. It leaves behind a negative charge, increasing the negative charge on the membrane side of the interface.
When a hydrogen ion bonds to an anion site in the glass, one or more molecules of water surrounding the ion are stripped away. At the same time, the charge separation decreases (Figure 3.10) because a charge on both sides of the interface has disappeared, one on the glass surface and one in solution at the outer Helmholtz plane (OHP). This mechanism accounts for the selectivity of the glass for H+ions as well as the variation in charge as a function of H+ion activity.
= Silicon = Oxygen = Hydrogen atom + = Hydrogen ion = Water dipole
+
OHP
+
+
OHP
+
+
– –
–
–
– –
FIGURE 3.10 Hydrogen ion transfer across the phase boundary from the outer Helmholtz plane (OHP) on the solution side to the glass surface. The negatively charged oxygen atom becomes a silanol (Si–O–H) group. With each ion that moves from the solution to the surface, the charge on both sides of the interface decreases by one.
k k
+ Pt
A B
Pt Ag –
FIGURE 3.11 Haugaard’s demonstration of sodium conductivity in glass. Beaker A contained a solu- tion of 0.02 M HCl and a glass bulb made of pH sensitive glass. The inner tube was filled with the same solution. Platinum electrodes were used on both the inside and outside of the bulb. Beaker B contained 0.02 M silver nitrate solution. A large voltage was applied between the platinum electrode inside the glass bulb in beaker A and the silver electrode in beaker B. As current passed, silver ions were reduced and plated out at the silver electrode in proportion to the total charge transferred through the system.
Source: Adapted with permission from Haugaard [20]. Copyright 1941, American Chemical Society.
Unlike the liquid membrane system in which K+ ions can diffuse through the membrane, H+ ions do not reach the other wall of the glass. Instead of hydrogen ions, sodium ions are responsible for the modest current. An elegant demonstration of sodium ion migration was reported by the Danish chemist, G. Haugaard in the 1940s [20]. He set up the apparatus depicted in Figure 3.11. A pH-sensitive glass bulb was filled with a dilute solution of HCl and placed into a beaker of the same solution. Haugaard placed a second container with a solution of silver nitrate next to the first one (Figure 3.11). A platinum electrode was used to bridge between the two beakers and a silver electrode dipping into the silver nitrate solution was connected to the negative side of the circuit.
Then, over a three-week period, a current was forced through the bulb (and the solution in the two beakers) using a large applied voltage (220 V).
At the end of the experiment, solution from inside beaker A was removed and evapo- rated to dryness. HCl is volatile, so if it were the only electrolyte in solution, there would be no residue. However, after evaporation a solid residue of NaCl was found in beaker A where no sodium was present previously. Haugaard was able to collect a weighable mass of sodium chloride from the solution on the negative side of the glass bulb and deter- mine the moles of sodium chloride after evaporation. In addition, for every charge passing through the system, an atom of silver was deposited on the silver wire.
Ag+(aq)+e⇌Ag(s) (3.21)
The mass of the silver wire was compared before and after the experiment. Haugaard demonstrated that the number of moles of sodium chloride accumulating in the solution outside the glass bulb was equal to the number of moles of silver gained by the silver wire.
k k The glass membrane was the only source for the sodium found in the outer solution. These
observations were a strong support for the conclusion that the current was being carried through the glass bulb by sodium ions.
Empirically, in experiments in which glass membranes were used to monitor hydro- gen ion activity, the voltage across the glass membrane was shown to follow a logarithmic dependence on the hydrogen ion activity. Both the solution/glass interface at the inside wall facing the reference solution and the sample solution/glass interface on the outer wall contribute to the membrane potential. As with conventional liquid membrane ISEs discussed earlier, the conditions for the reference side of the glass membrane are held constant, normally at pH 7.00, so that any measured potential changes can be assumed to be associated only with the outside interface between the sample solution and glass surface [15].